EP0280313B1

EP0280313B1 - Method of compressing image signals by vector quantization

Info

Publication number: EP0280313B1
Application number: EP88102850A
Authority: EP
Inventors: Nobuyuki C/O Fuji Photo Film Co. Ltd. Tanaka
Original assignee: Fuji Photo Film Co Ltd
Current assignee: Fujifilm Holdings Corp
Priority date: 1987-02-25
Filing date: 1988-02-25
Publication date: 1994-05-18
Anticipated expiration: 2008-02-25
Also published as: US4862262A; DE3889560T2; DE3889560D1; EP0280313A3; EP0280313A2

Description

This invention relates to a method of compressing image signals in accordance with the preamble of claims 1 and 4. This invention particularly relates to a method of compressing image signals so that a high signal compressibility is obtained by utilizing vector quantization and encoding by orthogonal transformation or prediction encoding in combination with each other.
A vector quantizer is already known from EP-A-0 097 858 which receives a plurality of input signal in block form and operates to determine an output vector having the least distortion with respect to an input vector by various techniques. A coded output or output vector codetable address is output from a coder of the device for receipt by a decoder which reconstructs the output vector signal.
Image signals representing half tone images, such as television signals, are composed of enormous amounts of information, and a broad-band transmission line is required for transmission of the image signals. Such image signals involve much redundancy, and various attempts have been made to compress the image signals by restricting the redundancy. Also, in recent years, recording of half tone images on optical disks, magnetic disks, or the like has been generally put into practice. In this case, image signal compression is conducted generally for the purpose of efficiently recording image signals on a recording medium.
One of the methods of image signal compression that has heretofore been known is a method wherein vector quantization is utilized. The known method comprises the steps of (i) dividing two-dimensional image signals into blocks each comprising the signals at MxN number of picture elements adjacent to one another, (ii) selecting a vector that corresponds with the minimum distortion to the set of the image signals in each of the blocks from a code book comprising a plurality of vectors different from one another and prepared in advance by defining MxN number of vector elements, and (iii) encoding the information representing the selected vector to correspond to the block.
Since the image signals in the block as mentioned above have high correlation therebetween, the image signals in each block can be represented very accurately by one of a comparatively small number of vectors prepared in advance. Therefore, transmission or recording of the image signals can be carried out by transmitting or recording a code representing the vector, instead of the image signals themselves, and signal compression can thus be achieved. By way of example, the amount of the image signals at 64 picture elements in a half tone image of a density scale composed of 256 levels (= 8 bits) is 8x64 = 512 bits. In the case where the 64 picture elements are grouped as a single block, the respective image signals in the block are expressed by a vector composed of 64 vector elements, and a code book including 256 such vectors is prepared, the amount of the signals per block becomes equal to the amount of the signals for discrimination between the vectors, i.e. 8 bits. Consequently, in this case, the amount of the signals can be compressed to 1/64.
After the image signals are compressed in the manner as mentioned above and recorded or transmitted in the compressed form, the vector elements of each of the vectors which the vector discriminating information represents are taken as reconstructed signals of each of the blocks, and the original image is reproduced by use of the reconstructed signals.
The aforesaid method of compressing image signals by vector quantization is advantageous for transmission of television signals and for other purposes. On the other hand, in recent years, an electronic image file for recording the images of very high gradation such as medical radiation images on the aforesaid optical disk or the like has attracted particular attention. For such purposes, a need exists for more efficient compression of the image signals.
The primary object of the present invention is to provide a method of compressing image signals, which improves the signal compressibility over the conventional method by utilizing the vector quantization technique.
Another object of the present invention is to provide a method of compressing image signals by vector quantization, which enables an increase in the number of images recordable on a recording medium or an increase in the signal transmission efficiency.
These objects are met by the features of claims 1 and 4.
With the first and second methods of compressing image signals by vector quantization in accordance with the present invention, signal compression is achieved to an extent higher than with the conventional method of compressing image signals utilizing vector quantization. Therefore, particularly in the case where images such as high-gradation medical images are to be recorded, the number of the images recordable on a recording medium can be increased markedly. Also, in the case where the first or second method of compressing image signals by vector quantization in accordance with the present invention is applied to image signal transmission, the signal transmission path can be reduced markedly, or the transmission time can be shortened markedly.
In general, the aforesaid encoding by orthogonal transformation is applied to the original image signals carrying a two-dimensional image for the purposes of signal compression. In this case, with the encoding by orthogonal transformation, the two-dimensional image signals are divided into blocks each of which is composed of an appropriate number of samples, and orthogonal transformation of a value string comprising the sample values is carried out for each block. Since energy is concentrated at a specific component by the orthogonal transformation, a component of high energy level is encoded (quantized) by allocating a long code length thereto, and a component of low energy level is encoded coarsely with a short code length, thereby reducing the number of codes per block. As the orthogonal transformation, Fourier transformation, cosine transformation, Hadamard transformation, Karhunen-Loeve transformation, or Harr transformation is usually used. The aforesaid image signal compression method will hereinbelow be described in further detail by taking the Hadamard transformation as an example. As shown in Figure 2, blocks are formed by dividing digital two-dimensional image signals in a unit of, for example, two signals in a predetermined one-dimensional direction. When sample values x(0) and x(1) in the block are plotted on an orthogonal coordinate system, since correlation therebetween is high as mentioned above, most of the sample values are distributed near the straight line represented by the formula x(1 ) =x(0) as shown in Figure 3. Therefore, as shown in Figure 3, the orthogonal coordinate system is transformed by an angle of 45 _°to determine a new y(0)-y(1) coordinate system. On the y(0)-y(1) coordinate system, y(0) represents the low frequency component of the original image signals prior to transformation, and attains a value slightly larger than x(0) and x(1) [a value approximately .j2 times the values of x(0) and x(1)]. On the other hand, values of y(1) representing the high frequency component of the original image signals are distributed just within a very narrow range near the y(0) axis. In the case where a code length of, for example, seven bits is required for encoding of x(0) and x(1), seven bits or eight bits are required for encoding of y(0). On the other hand, y(1) can be encoded with a code length of, for example, four bits. Consequently, the code length per block is shortened, and compression of the image signals is achieved.
Orthogonal transformation of second order wherein each block is constituted by two image signals is conducted as mentioned above. As the order is increased, the tendency of energy concentrating at a specific component is increased, and it becomes possible to improve the effect of decreasing the number of bits. In general, the aforesaid transformation can be conducted by use of an orthogonal function matrix. In an extreme case, when an intrinsic function of the objective image is selected as the orthogonal function matrix, the transformed image signals are constituted by the intrinsic value matrix, and the original image can be expressed just by the diagonal component of the matrix. Also, instead of dividing the image signals just in a single direction to form blocks as mentioned above, each block may be constituted by two-dimensional image signals. In this case, the effect of decreasing the number of bits is markedly improved over the one-dimensional orthogonal transformation.
The transformed signals obtained by the aforesaid two-dimensional orthogonal transformation are arranged in the order of the sequency (i.e. the number of "0" position crossing) of the orthogonal function utilized for the transformation in each block. Since the sequency is correlated with the spatial frequency, the transformed signals are put side by side in the longitudinal and transverse directions in the order of the frequency as shown in Figure 4. Therefore, the code length per block may be shortened by allocating a comparatively long code length to the transformed signals representing the low frequency component (i.e. the signals on the left upper side in Figure 4) as in the case where a longer code length is allocated to y(0) in the aforesaid one-dimensional orthogonal transformation of second order, and allocating a comparatively short code length or no code to the transformed signals representing the high frequency component (i.e. the signals on the right lower side in Figure 4).
In the first method of compressing image signals by vector quantization in accordance with the present invention, instead of applying the aforesaid encoding by orthogonal transformation to the original image signals, the encoding by orthogonal transformation is applied to the signals of the numbers representing the vectors selected by vector quantization. In this case, signal compressing can be achieved by defining the vector numbers in the manner as mentioned above. Specifically, in the case where the vector numbers are defined in the manner as mentioned above, the numbers of the vectors become closer to one another as the defined vector distortions are smaller. On the other hand, the original image signals have high correlation among the picture elements close to one another in the block B1. Also, in the case where a plurality of the blocks B1 are considered, the correlation of the overall image signals becomes high among the blocks B1 that are close to one another. Therefore, the correlation among the vectors selected for the respective blocks B1 becomes high among the blocks B1 that are close to one another, and consequently the correlation among the numbers of the vectors becomes high. Accordingly, when the vector numbers are subjected to the encoding by orthogonal transformation, the code lengths necessary for representing the vector numbers can be reduced.
Also, the aforesaid prediction encoding is applied to the original image signals carrying a two-dimensional image for the purposes of signal compression. Image signals (remark image signals) have high correlation with image signals in the vicinity of the remark image signals. Therefore, when the values of the remark image signals are predicted from the values of the image signals in the vicinity of the remark image signals, the distribution of the prediction errors is concentrated in the vicinity of zero. In the aforesaid prediction encoding, by the utilization of the characteristics that the prediction errors are concentrated in the vicinity of zero, instead of encoding the actual image signals, the prediction errors which can be expressed with short codes are encoded. In this manner, the total amount of signals is compressed.
In the second method of compressing image signals by vector quantization in accordance with the present invention, instead of applying the aforesaid prediction encoding to the original image signals, the prediction encoding is applied to the signals of the numbers representing the vectors selected by vector quantization. In this case, signal compressing can be achieved by defining the vector numbers in the manner as mentioned above. Specifically, in the case where the vector numbers are defined in the manner as mentioned above, the numbers of the vectors become closer to one another as the defined vector distortions are smaller. On the other hand, the original image signals have high correlation among the picture elements close to one another in the block. Also, in the case where a plurality of the blocks are considered, the correlation of the overall image signals becomes high among the blocks that are close to one another. Therefore, the correlation among the vectors selected for the respective blocks becomes high among the blocks that are close to one another, and consequently the correlation among the numbers of the vectors becomes high. Accordingly, when the vector numbers are subjected to the prediction encoding, the code lengths necessary for representing the vector numbers can be reduced.

Figure 1 is a block diagram showing the configuration of an apparatus for carrying out the first method of compressing image signals by vector quantization in accordance with the present invention,
Figures 2, 3 and 4 are explanatory views showing the orthogonal transformation in the first method of compressing image signals by vector quantization in accordance with the present invention,
Figure 5 is an explanatory view showing the division of image signals into blocks,
Figure 6 is an explanatory view showing the division of vector discrimination signals into blocks in the first method of compressing image signals by vector quantization in accordance with the present invention,
Figure 7 is a block diagram showing the configuration of an apparatus for carrying out the second method of compressing image signals by vector quantization in accordance with the present invention,
Figure 8 is a schematic view showing examples of vector discrimination signals in the second method of compressing image signals by vector quantization in accordance with the present invention, and
Figures 9A, 9B and 9C are explanatory views showing the prediction encoding of the vector discrimination signals in the second method of compressing image signals by vector quantization in accordance with the present invention.

The present invention will hereinbelow be described in further detail with reference to the accompanying drawings.
With reference to Figure 1, original image signals S representing a single continuous tone image are fed to a block transformation circuit 10 in which they are transformed into original image signals x of each of rectangular blocks which comprises MxN picture elements. The division into the blocks is shown in Figure 5. With reference to Figure 5, F denotes the original image, and B1 denotes a single block. For simplicity of explanation, the block B1 is assumed to comprise 6x6 picture elements in the descriptions below.
The original image signals x of the block B1 are then fed to a vector quantizer 11. The vector quantizer 11 selects a vector, that corresponds with the minimum distortion to the set of the fed original image signals x of the block B1 (the set comprises 36 signals), from a plurality of vectors stored as a code book in a memory 12. Specifically, the memory 12 stores a code book representing, by way of example, 256 vectors 'x'(1),'x'(2),'x'(3),..., X (256)respectively defining 36 vector elements
where n = 1, 2, ..., 256 as shown below.
The vector quantizer 11 finds a vector x(m)whose vector elements (x̂₁, x̂₂, x̂₃, ..., x̂₃₆) correspond with the minimum distortion to the set (x₁, x₂, x₃, ..., x₃₆) of the original image signals x, and outputs an encoded signal Dm which represents a vector identification number "m" representing the vector x (m). As the distortion, by way of examples, the mean square error expressed as
is utilized (k = 36 in this example). In order to find such a vector x (m)that the distortion is the minimum, the distortion may be calculated for all of the vectors, and the vector x (m)exhibiting the minimum distortion may be found (this method is referred to as total search type vector quantization). Alternatively, in order to shorten the processing time, binary tree search type vector quantization may be carried out though the distortion may not completely become the minimum.
In this embodiment, the vectors x (2), x (3),..., x (256)after the vector x (1) are put side by side in such a sequence that the distortion with respect to the vector x (1) is smaller. Specifically, by way of example, the distortion between the vector x (m) of the identification number "m" and the vector x (m + 1) of the identification number (m + 1) is smaller than the distortion between the vector x (m) and the vector x (m + 2) of the identification number (m+2). The distortion may be defined by, for example, the aforesaid mean square error.
An appropriate code book comprising the vectors respectively defining the vector elements (x̂₁ x̂₂, x̂₃, ..., x̂₃₆) may be prepared by preparing a training image of the same type as the image on which signal compression is to be carried out, and using a known method on the basis of the training image.
In this embodiment, the vector discrimination signal Dm may be such that the 256 vectors can be discriminated from one another, and can therefore be expressed by 8 bits. Therefore, in the case where the density scale of each picture element is of 256 levels (= 8 bits), the image signals in the amount of 8 bits x 36 (picture elements) can be expressed by 8 bits, and the signal compressibility becomes 1/36.
The vector selection and the output of the vector discrimination signal Dm as mentioned above are carried out for all of the blocks B1 in a single image which the original image signals S represent. The vector identification numbers which the vector discrimination signals thus generated represent are distributed as shown in, for example, Figure 6 for a single image. Specifically, in this embodiment, a single vector identification number is determined for 6x6 = 36 picture elements. Also, as mentioned above, the vector identification numbers have higher correlation among the blocks B1 that are closer to one another. As shown in Figure 1, the vector discrimination signals Dm representing the vector identification numbers are fed to an orthogonal transformation circuit 13 and subjected therein to, for example, two-dimensional orthogonal transformation. As shown in Figure 6, the two-dimensional orthogonal transformation is carried out for each of blocks B2 each of which is composed of PxQ number of the blocks B1 adjacent to one another. By way of example, in Figure 6, 3x3 = 9 blocks B1 are taken as a single block B2.
By way of example, as the orthogonal transformation, Hadamard transformation as mentioned above may be used. Since the transformation matrix in Hadamard transformation is constituted just by + 1 and -1, Hadamard transformation can be executed by use of a transformation circuit simpler than in the other orthogonal transformations. Also, as well known, two-dimensional orthogonal transformation can be reduced to one-dimensional orthogonal transformation. Specifically, the two-dimensional orthogonal transformation is carried out by subjecting the PxQ number of the vector discrimination signals Dm in the two-dimensional block B2 to one-dimensional orthogonal transformation in the longitudinal direction, and then subjecting the PxQ number of transformed signals thus obtained to one-dimensional orthogonal transformation in the transverse direction. The transformation in the longitudinal direction and the transformation in the transverse direction may be carried out in the reverse order.
The transformed signals Ym obtained by the two-dimensional orthogonal transformation are put side by side in each block B2 in the longitudinal and transverse directions in the order of the sequency of the function on which the orthogonal transformation is based (for example, the Walsh function in the case of Hadamard transformation, or the trigonometrical function in the case of Fourier transformation).
The transformed signals Ym put side by side in the order of the sequency of the function on which the two-dimensional orthogonal transformation is based are sent to a known encoding circuit 14 as shown in Figure 1, and encoded (quantized) therein. As mentioned above, energy is concentrated at a specific component in the transformed signals Ym. Therefore, it is possible to decrease the number of bits required per block B2 and to achieve signal compression by allocating a comparatively long code length to the component with high energy, and allocating a comparatively short code length to or discarding the component with low energy.
Signals f(m) representing the vector identification numbers encoded in the manner as mentioned above are recorded on a recording medium (image file) such as an optical disk or a magnetic disk in a recording and reproducing apparatus 15. As mentioned above, the signals f(m) have been markedly compressed by vector quantization and encoding by orthogonal transformation as compared with the original image signals S. Therefore, a large number of the images can be recorded on the recording medium such as the optical disk.
In the course of image reproduction, the signals f(m) are read from the recording medium, and are decoded by a decoder 16 into the aforesaid transformed signals Ym. The transformed signals Ym thus decoded are sent to an inverse transformation circuit 17 and transformed inversely to the aforesaid two-dimensional orthogonal transformation. In this manner, the vector discrimination signals Dm representing the vector identification numbers are restored. The vector discrimination signals Dm indirectly representing the image signals are transformed by a decoder 18 into reconstructed signals x'. Specifically, the decoder 18 reads the vector, which the vector discrimination signal Dm fed thereto represents, from the code book stored in the memory 12, and outputs the vector elements (X_l, )(2, )(3, ..., X₃₆), which are defined for said vector, as the reconstructed signals x' for a single block B1.
The reconstructed signals x' are sent to a composing circuit 19 and transformed therein from the signals per block B1 to the signals per image. The image signals S' obtained by said transformation in the composing circuit 19 have slight distortion with respect to the original image signals S, and are approximately equal to the original image signals S. The image signals S' are ultimately sent to an image reproducing apparatus 20. In the image reproducing apparatus 20, an image approximately identical with the original image that the original image signals S represent is reproduced on the basis of the image signals S'.
An embodiment of the second method of compressing image signals by vector quantization in accordance with the present invention will hereinbelow be described with reference to Figure 7.
In this embodiment, the original image signals S representing a single continuous tone image are fed to a block transformation circuit 10 in which they are transformed into original image signals x of each of rectangular blocks B which comprises MxN picture elements as shown in Figure 5. For simplicity of explanation, each of the blocks B is assumed to comprise 6x6 picture elements in the descriptions below.
The original image signals x of the block B are then fed to the vector quantizer 11. In the same manner as in the embodiment shown in Figure 1, the vector quantizer 11 selects the vector x(m),that corresponds with the minimum distortion to the set of the fed original image signals x of the block B1 (the set comprises 36 signals), from a plurality of vectors stored as the code book in the memory 12, and outputs the encoded signal Dm which represents the vector identification number "m" representing the vector x (m).
The vector selection and the output of the vector discrimination signal Dm as mentioned above are carried out for all of the blocks B in a single image which the original image signals S represent. The vector identification numbers which the vector discrimination signals thus generated represent are distributed as shown in, for example, Figure 8 for a single image. Specifically, in this embodiment, a single vector identification number is determined for 6x6 = 36 picture elements. Also, as mentioned above, the vector identification numbers have higher correlation among the blocks B that are closer to one another.
As shown in Figure 7, the vector discrimination signals Dm representing the vector identification numbers are fed to a prediction encoding circuit 23 and subjected therein to previous value prediction plus Huffman encoding processing as an example of the prediction encoding processing methods. For explanation of prediction encoding processing, the group signals of the vector identification numbers as shown in Figure 8 are generally indicated as shown in Figure 9A. First the prediction encoding circuit 23 carries out previous value prediction for the aforesaid signals a11, a12, a13, ..., a21, a22, a23, ..., a31, a32, a33, ..., and calculates prediction errors Δa12, Δa13, ..., Aa22, Aa23, ..., Aa32, Aa33, ..., as shown in Figure 9B. The image signals standing at the head of lines L1, L2, L3, ... are left as they are. With previous value prediction, the value of a remark signal is predicted to be equal to the value of the signal present prior to the remark signal, and the difference (i.e. the prediction error) between the predicted value (i.e. the value of the signal present prior to the remark signal) and the actual value of the remark signal is calculated. For example, the prediction errors are calculated as
After calculating the prediction errors in the manner as mentioned above, the prediction encoding circuit 23 encodes the raw signals a11, a21, a31, ... standing at the head of the lines L1, L2, L3, ... and the prediction errors Δa12, Δa13, ..., Δa22, Δa23, ..., Aa32, Δa33, ... on the basis of a Huffman code table, and generates compressed signals a11', Δa12', Δa13', ... as shown in Figure 9C. As mentioned above, the vector discrimination signals Dm representing the vector identification numbers can be compressed by carrying out prediction encoding in this manner.
Signals f(m) representing the vector identification numbers prediction-encoded in the manner as mentioned above are recorded on a recording medium (image file) such as an optical disk or a magnetic disk in the recording and reproducing apparatus 15. As mentioned above, the signals f(m) have been markedly compressed by vector quantization and prediction encoding as compared with the original image signals S. Therefore, a large number of the images can be recorded on the recording medium such as the optical disk.
In the course of image reproduction, the signals f(m) are read from the recording medium, and are decoded by a decoder 27 into the aforesaid vector discrimination signals Dm representing the vector identification numbers. Specifically, with decoding processing, the compressed signals as shown in Figure 9C are decoded based on the Huffman code table which was used in the course of effecting the aforesaid Huffman encoding processing. In this manner, the raw signals at the head of the respective lines and the prediction errors as shown in Figure 9B are obtained. Then, by use of the raw signals at the head and the prediction errors obtained in this manner and the prediction formulas used in the course of the aforesaid previous value prediction, the extended signals a11, a12, a13, ..., a21, a22, a23, ..., as shown in Figure 9A are calculated as shown below.

The decoder 27 outputs the vector discrimination signals Dm (a group of the signals a11, a12, a13, ..., a21, a22, a23, ...) decoded in the manner as mentioned above.
The vector discrimination signals Dm indirectly representing the image signals are transformed by the decoder 18 into reconstructed signals x'. Specifically, the decoder 18 reads the vector, which the vector discrimination signal Dm fed thereto represents, from the code book stored in the memory 12, and outputs the vector elements (x̂₁, x̂₂, x̂₃, ..., x̂₃₆), which are defined for said vector, as the reconstructed signals x' for a single block B.
The reconstructed signals x' are sent to the composing circuit 19 and transformed therein from the signals per block B to the signals per image. The image signals S' obtained by said transformation in the composing circuit 19 have slight distortion with respect to the original image signals S, and are approximately equal to the original image signals S. The image signals S' are ultimately sent to the image reproducing apparatus 20. In the image reproducing apparatus 20, an image approximately identical with the original image that the original image signals S represent is reproduced on the basis of the image signals S'.
In the embodiments shown in Figures 1 and 7, the vector identification numbers 2, 3, 4, ..., 256 are assigned to the vectors put side by side in such a sequence that the distortion with respect to the vector x (1) is smaller. Instead, the identification number 256 may be assigned to a predetermined vector x, thereby to indicate it as x (256), and the vector identification numbers 255, 254, 253, ..., 3, 2, 1 may be assigned to the vectors put side by side in such a sequence that the distortion with respect to the vector x (256) is smaller. Also, the predetermined vector which is taken for reference in assignment of the identification numbers to the respective vectors x and putting them side by side may be a vector that is not defined in the code book.

Claims

1. A method of compressing image signals by vector quantization, which comprises the steps of:

i) dividing two-dimensional image signals (S) into input blocks (B1) each of which is composed of a number of picture elements (M^*N) adjacent to one another,

ii) for each of said input blocks (B1), selecting an output vector that corresponds to the input block (B1) with minimum distortion from a code book (12) comprising a plurality of output vectors different from one another and prepared in advance, and

iii) producing a quantized signal (Dm) representing said selected output vector to correspond to each of the input blocks (B1),

characterized by the steps of:

a) for each of said output vectors in said code book (12), defining a number (m) to be used as said signal (Dm) so that said number (m) for a given output vector is a monotonically increasing or decreasing function of the distortion (d) of the given output vector in said code book (12) with respect to a predetermined vector,

b) grouping said quantized signals (Dm) selected for the respective input bolcks (B1) into blocks (B2), each of which is composed of a number of quantized signals (P^*Q) adjacent to one another,

c) carrying out an orthogonal transformation (13) of said quantized signals in each of said blocks (B2), and

d) encoding (14) the transformed signals (Ym) obtained by said orthogonal transformation (13) with an intrinsic code length.

2. A method as defined in claim 1, wherein said orthogonal transformation (13) is a two-dimensional orthogonal transformation.

3. A method as defined in claim 1 or 2, wherein said orthogonal transformation (13) is a Hadamard transformation.

4. A method of compressing image signals by vector quantization, which comprises the steps of:

i) dividing two-dimensionsl image signals (S) into input blocks (B) each of which is composed of a number of picture elements (M^*N) adjacent to one another,

ii) for each of said input blocks (B1), selecting an output vector that corresponds to the input block (B) with minimum distortion from said code book (12) comprising a plurality of vectors different from one another and prepared in advance, and

iii) producing a quantized signal (Dm) representing said selected output vector to correspond to each of the input blocks (B),

characterized by the steps of:

b) carrying out prediction encoding (23) of said quantized signals of said numbers selected for the respective blocks (B).

5. A method as defined in claim 4, wherein said prediction encoding (23) is previous value prediction plus Huffman encoding processing.